Title: How a Single Paper Affects the Impact Factor: Implications for Scholarly Publishing

Abstract: Because the Impact Factor (IF) is an average quantity and most journals are small, IFs are volatile. We study how a single paper affects the IF using data from 11639 journals in the 2017 Journal Citation Reports. We define as volatility the IF gain (or loss) caused by a single paper, and this is inversely proportional to journal size. We find high volatilities for hundreds of journals annually due to their top-cited paper: whether it is a highly-cited paper in a small journal, or a moderately (or even low) cited paper in a small and low-cited journal. For example, 1218 journals had their most cited paper boost their IF by more than 20%, while for 231 journals the boost exceeded 50%. We find that small journals are rewarded much more than large journals for publishing a highly-cited paper, and are also penalized more for publishing a low-cited paper, especially if they have a high IF. This produces a strong incentive for prestigious, high-IF journals to stay small, to remain competitive in IF rankings. We discuss the implications for breakthrough papers to appear in prestigious journals. We also question the practice of ranking journals by IF given this uneven reward mechanism.